The Intergenerational Transmission of Poverty during the AIDS Epidemic … · 2016-08-02 · He has...

Working Paper October 2009 No. 100

www.chronicpoverty.org Chronic Poverty Research Centre

ISBN: 1-904049-99-0

What is Chronic Poverty?

The distinguishing feature of chronic poverty is extended duration in absolute poverty.

Therefore, chronically poor people always, or usually, live below a poverty line, which is normally defined in terms of a money indicator (e.g. consumption, income, etc.), but could also be defined in terms of wider or subjective aspects of deprivation.

This is different from the transitorily poor, who move in and out of poverty, or only occasionally fall below the poverty line.

Assets and chronic poverty:

background paper

Andrew McKay

University of Sussex Falmer, Brighton, BN1 9RH United Kingdom

Assets and chronic poverty: background paper

2

Abstract

The role of assets is key to the study of changes in welfare outcomes. Assets can play an

important role in reducing vulnerability, which is a key dimension of both chronic and

transient poverty. Assets are also important in influencing what households are able to

achieve, in terms of income and many other outcomes. Those with more assets are often

better able to improve their income levels and so to participate in economic growth.

Accumulation of assets is a means by which people can move out of poverty and improve

their livelihoods. On the other hand, those losing assets may be pushed into poverty. And

those lacking assets to begin with risk being caught in a poverty trap. This paper begins by

discussing the relationship between assets (or their absence) and vulnerability, given the

central role of vulnerability in understanding both chronic and transient poverty. It then

discusses the role of changes in asset holdings in poverty transitions. This leads into a

discussion of asset-based poverty traps and of the potential role of discrimination. It

concludes by setting out some areas for further research.

Keywords: assets, vulnerability, poverty traps, chronic poverty

Acknowledgements

I am grateful to Martin Prowse for very helpful, extensive and detailed research assistance,

relating in particular to Section 3 of this paper.

Andrew McKay is Economics Professor at the University of Sussex and is currently head of

department. He has been associated with the Chronic Poverty Research Centre since its

inception, and has been an associate director since 2005. He has extensive field experience

in Africa, and has some experience in Asia, most recently in Vietnam.

Email: [email protected]

mailto:[email protected]


3

Contents

1 Introduction ..................................................................................................................... 4

2 Assets and vulnerability ................................................................................................. 6

3 Asset accumulation, loss and poverty transitions ........................................................ 9

3.1 Adverse asset shocks and descents into poverty ........................................................................ 9

3.2 Asset accumulation and escapes from poverty .......................................................................... 11

3.3 Returns to assets and poverty transitions .................................................................................. 12

3.4 Other factors in relation to assets ............................................................................................... 13

4 Assets in relation to poverty dynamics ........................................................................15

5 Discrimination as a factor in access to and use of assets ..........................................21

6 Conclusions ....................................................................................................................24

References ..........................................................................................................................26


4

1 Introduction

By definition, the Chronic Poverty Research Centre (CPRC) adopts a dynamic approach to

the study of poverty. This calls for a focus on changes in key welfare outcomes. But within

such framework, the role of assets is key. The assets which individuals own or have access

to matter in at least two respects. First, assets can play an important role in reducing

vulnerability, which is a key dimension of both chronic and transient poverty. There is

extensive evidence that assets help provide insurance against shocks, reducing insecurity

and frequently reducing risk-averse behaviour and reliance on more destructive coping

strategies – which commonly involve reducing asset levels, e.g. withdrawing children from

school. Of course many households, especially those in chronic poverty, may not have

access to sufficient assets, which then limits their ability to cope with vulnerability. Second,

assets play an important role in influencing what households are able to achieve, in terms of

income and many other outcomes. Those with more assets are often better able to improve

their income levels and so to participate in economic growth (for example, by having better

access to credit), as well as being better able to protect themselves against downturns.

Accumulation of assets is an important means by which people can move out of poverty and

improve their livelihoods. Having more assets also potentially plays an important role in

providing – and indicating – social status, and potentially benefiting more from public policy

interventions. On the other hand, those losing assets – perhaps as the result of a health

shock – may be pushed into poverty. And those lacking assets to begin with risk being

caught in a poverty trap (Carter and Barrett, 2006), or at least left behind in an environment

of positive change. There is an important risk of an inequality dynamic which can be

understood in terms of assets.

A wide range of assets is important in this, as suggested in a livelihoods framework. As well

as the level of assets to which individuals and households have access, it is also important to

consider the uses to which they can put these assets. A higher level of human capital is less

valuable if an individual is not able to obtain a skilled job; or if she receives a lower wage,

due to discrimination. The inability to use assets effectively (‘low returns’ to assets) may

contribute to chronic poverty; while the ability to earn an adequate return on assets may be

an important factor underlying escapes from poverty – perhaps over a period of time. The

factors which affect how individuals are able to use their assets are very important. In

addition, the need of some households to sell assets is also a key potential cause of chronic

poverty if they have few assets to begin with. Understanding why some need to make these

choices, while others can avoid them, is important in understanding chronic poverty.

Assets are also of interest as a potential indicator of poverty in their own right, although there

is a challenge in trying to combine information on many different assets. Many individual or

household-level assets tend to fluctuate much less over time than income or often


5

consumption, and as such might be better indicators of longer-term welfare prospects. Lack

of key assets therefore may be a good proxy for chronic poverty.

The pattern of access to assets – whether through ownership or not – and of the uses to

which these assets can be put is therefore of key importance in analysing vulnerability,

poverty dynamics and chronic poverty. There are key political dimensions underlying these

issues. It is important to understand the processes through which some individuals (or

households) are able to own and accumulate assets, while others are not; and why some are

able to use their assets more effectively than others.

While these issues have been discussed to date at the individual or household (micro) level,

analogous issues arise at meso and macro levels. The fact that an individual or household is

located in a lagging region of a country (in which other regions are growing) may itself be a

major factor underlying chronic poverty; alternatively, specific villages might be considered

as persistently poor due to specific disadvantages they face. Being poor in a country which is

experiencing sustained economic decline – or at least no consistent economic progress – is

likely to imply being in chronic poverty. The meso and macro contexts clearly matter. And

these meso and macro contexts can potentially also be considered in terms of assets,

returns to assets and vulnerability.

This paper begins by discussing the relationship between assets (or their absence) and

vulnerability, given the central role of vulnerability in understanding both chronic and

transient poverty. It then discusses in Section 3 the role of changes in asset holdings in

poverty transitions. This leads into a discussion of asset-based poverty traps (Section 4) and

of the potential role of discrimination (Section 5). Section 6 concludes, setting out some

areas for further research.


6

2 Assets and vulnerability

In practice, poor people often manage relatively complex asset portfolios. Poverty reduction

policy therefore should focus on what poor people have (assets) as much as what they lack

(Moser, 1998), and should seek to help them to accumulate assets and manage more

effectively what they have. Assets provide poor people with a key means of coping with

shocks. Equally, the erosion of assets translates into increased insecurity.

The importance of assets in relation to vulnerability has been extensively developed, in

particular in rural areas (Swift, 1989, Sen, 1991, Maxwell and Smith, 1992), and Barrett

(1999) highlights the key role that assets play in relation to food security. Having cash

savings means that food can be purchased; having land plus sufficient labour enables

cultivation; having a strong supporting network or access to government may allow access to

goods even without cash. There is much evidence of assets providing a buffer against

idiosyncratic and covariant risks and being an important determinant of social status (Barrett

et al., 2006a). But assets themselves are not sufficient; institutions and technology are also

important (Barrett, 1999), e.g. access to markets or production technology, or the

effectiveness of government or private food distribution.

Households cope with adverse shocks in different ways (Chambers, 2003; Dercon, 2000.

They often rely initially on various ex post mechanisms to try to smooth consumption levels,

e.g. diversifying activities, adjusting consumption or drawing down savings. Sometimes sales

of assets may be part of this response, though hopefully not to such an extent as to

compromise their future security. Households’ ability to respond in this way depends on the

assets they own, as well as on their market access. A combination of assets plus market

access can mean that variability arising from, say, climate or market variability need not

translate into physical vulnerability. Where market-based responses do not succeed,

households will typically seek to draw on the ‘moral economy’: support networks of family

and friends (Lourenço-Lindell, 2002; Woolcock, 2005).. The effectiveness of such

mechanisms depends on there being a sufficiently developed level of social capital, as well

as on the shocks not being covariate, such that the family and friends are equally affected.

Formal assistance programmes offer a third channel of support in the face of shocks

(Devereux, 2002).

In urban areas the environment, and sometimes the assets which are relevant, may be

significantly different (Moser, 1998; Mitlin, 2003; Nkurunziza and Rakodi, 2005). More

facilities may be available, but the extent of commoditisation will often be much greater than

rural areas; and in addition there may be important environmental hazards, such as poor

housing, inadequate water and sanitation, as well as pollution. The moral economy may be

less well developed than rural areas. The types of assets which are often important in urban

areas include labour and human capital; productive assets, among which housing may be

very important; as well as less tangible assets, such as household relations and social capital


7

(Moser, 1998). The majority of even poor households benefit from access to some or all of

these assets, and they form a basis of a household’s responses to shocks. There are often

significant adjustments in a household’s supply of labour. This may involve women or

children becoming more involved in work (which may nonetheless have adverse longer-term

consequences, for example for family relations; or a family member may migrate elsewhere

(with, again, potential risks for family relations). Housing is often used as an income-

generating asset through home-based economic activities. Household composition is

frequently flexible and may adjust as part of a response to a shock (e.g. migration or taking in

other family members). And social capital is often a key component of a response to a shock,

through, for example, more community-based activities or increased reliance on social

support networks (González de la Rocha, 2006; 2007; Lourenço-Lindell, 2002).

In responding to shocks households will often seek to protect the assets they have, and this

may then involve modifying consumption. Devereux (1993) considers consumption-modifying

strategies in response to shocks (see also Corbett, 1988); and Moser (1998) focuses some

of her discussion on household responses to consumption which do not damage their net

asset position. In terms of empirical evidence, Fafchamps et al. (1998) find that, in a period

of severe droughts in Burkina Faso in the early 1980s, only a small proportion of the income

shortfall was made up by households selling their livestock. Nor was there any evidence that

this shortfall was compensated by food stocks, off-farm work or transfers. The benefit of

holding onto livestock may have been a high return after the drought ceased; it may also be

that livestock are not a liquid asset in Burkina. But it seems that some of the adjustment to

the shocks was born by households adjusting their consumption levels. There is evidence

from Uganda too of households adjusting their consumption in response to shocks

(reference).

It is important to remember though that poor people, by the very fact of their poverty, have a

limited asset base. They are often more highly reliant on natural resources (Arun, 2008),

which may be common property resources, than the non-poor, and this potentially exposes

them to greater risks. They will typically receive inadequate protection from the law, may be

at greater risk from possible conflict, and may often be discriminated against. If, in addition,

they lack voice (as is frequently the case), then they are unlikely to be able to demand

access to assets and to protect themselves under the law and against discrimination. The

asset poverty of the poor reflects to a substantial extent their adverse incorporation in society

(Arun, 2008). Inappropriate asset management practices (e.g. excessive exploitation of

common property resources or shortened fallow cycles in cultivation) may further enhance

their vulnerability. In many countries, the fact that women may not have their right of access

to land legally recognised is a severe constraint in their ability to use this asset effectively in

reducing vulnerability. Thus, while poor people do have assets – and that is an important

resource to build on in designing policy interventions – the limited level of these assets and

households’ difficulty in managing them effectively (the latter linked to institutional,


8

governance and discrimination issues, among other things) severely constrains their ability to

use assets to reduce vulnerability.

Social capital potentially plays a very key role. It was stressed above as a key coping

mechanism, and the limited level of assets and difficulty in managing them emphasises this

more. In general, social capital is likely to be more important for poor people (though they

might have lower levels) than for non-poor people, but political capital also matters (Arun,

2008). Development of social capital implies, among other things, development of effective

community-based organisations (CBOs), and Moser (1998) reports that horizontal CBOs

were quite well developed in urban squatter settlements she studied. But it is also important

to recognise the limits of social capital. Moser (1998) reports that in the urban areas she

studies people often fail to support their communities once their other assets are depleted.

Social capital may be important, but is potentially of limited use without, for example,

housing, friends or education. Investment in social capital is therefore not a substitute for

investment in other forms of assets for poor people. Investment in these other assets (for

instance, human capital, providing tenure security for land and housing, or providing credit to

squatters) is critical in order to reduce the vulnerability that poor people face.

Assets of all types do play a key role in reducing the vulnerability that households face. But

the limited level of assets that many households have, plus severe constraints in being able

to manage these effectively, is a major contributor to high levels of vulnerability, as well as

persistent poverty.


9

3 Asset accumulation, loss and poverty transitions

There is now a lot of empirical evidence, much based on panel data, in support of the role of

asset loss (often a sudden phenomenon, as in a major health shock) or asset accumulation

(often a gradual process) being a correlate – and probably a cause – of descents into poverty

or escapes from poverty, respectively. There is also a wide literature on how the ways

individuals and households are able to use their assets influence their livelihood trajectories

and poverty levels. Each of these issues is briefly commented on in turn here, drawing on

selected recent empirical literature.

3.1 Adverse asset shocks and descents into poverty

As argued above, there is extensive evidence of the importance of loss of key assets as

major factors associated with descents into poverty – poverty which is potentially long term in

many instances. Much of this relates to health shocks, health being a key dimension of

human capital (as well as an important welfare indicator in its own right). Many empirical

studies demonstrate this (Harpham and Grant, 2002).. Thus in rural Uganda, Krishna et al.

(2006) found factors associated with illness as being the primary cause of descents into

poverty, with various factors being important, including the direct effect of ill health itself, high

healthcare expenses and the death of an income earner. Land-related factors – including the

impact of land division (associated with inheritance) and land exhaustion – were also strongly

associated with descents into poverty.

Ahmed (2005) finds that the cost of healthcare limits its use by the poorest in Bangladesh,

resulting in increased morbidity rates and destructive coping strategies (e.g. sale of assets or

accessing credit on highly unfavourable terms), which result in a deeper descent into

poverty. Krishna (2004) also report evidence from Rajasthan, India, of health shocks leading

to entries into poverty and increased levels of indebtedness, with the extortionate interest

rates charged on such loans being a major factor keeping households poor. These examples

highlight the importance of credit market failures in keeping people poor.

Small household size is found to make the impact of health shocks more severe in North

East Ghana, where a vicious circle is identified between poverty and small household size

(Whitehead, 2006). Small households are much less able to resist the impact of illness and

mortality shocks. Morbidity tended to lead to adverse impacts on food production, but larger

households were more able to cope with these shocks. This is in contrast to the more

common assertion that poverty levels are higher in large households, though is consistent

with some other findings (e.g. Howe and McKay, 2007, on chronically poor people in

Rwanda).


10

HIV/AIDS represents a severe adverse health shock. Besides its direct effects, it also has a

serious adverse affect on human capital formation through various channels: deaths of

teachers and trainers; reduced life expectancy reducing returns to investment in education;

parents potentially spreading education more widely among their children reducing the

likelihood that any one achieves the threshold level necessary to engage in high return

activities; increased health expenditures; reduced savings rates; as well as asset liquidation

and/or reduced consumption levels (Hamoudi and Birdsall, 2001).

Other adverse shocks affecting a household may also lead to reduced human capital

formation in education, and these effects are not adequately captured by simply looking at

the impact of conventional educational indicators such as enrolment rates. In Peru, Escobal

et al. (2005) report various adverse effects of shocks which do not necessarily show up in

enrolment rates, including reduced expenditure on education, reduced time devoted to

schooling, and movements from better quality private to poorer quality public schools.

In Bangladesh, Sen (2004) finds that various adverse assets shocks are associated with

descents into poverty, including life cycle changes (generally increased dependency rates);

health shocks again; and natural crises, such as flooding. In Kenya and Madagascar, Barrett

et al. (2006b) also highlight the importance of asset shocks in declines, in particular shocks

associated with health, land or labour. In Uganda, the belief of urban elites that descents into

poverty in rural areas are associated with drunkenness and laziness is not supported by the

findings of Krishna et al. (2006), except in a few specific cases.

It is quite clear that the loss of assets is a major factor associated with descent into poverty,

with the strongest evidence for this relating to diverse health shocks. The ability of

households to cope with these adverse shocks without employing destructive coping

strategies is clearly a major factor in enabling them to recover. Hoddinott (2006) finds in

Uganda that the children from more wealthy households, whose BMI was adversely affected

by a rainfall shock, were able to recover their growth trajectories afterwards, while those in

poorer households were not. Assembling more evidence on the factors which enable

households to cope from asset – or other – shocks without compromising their long-term

livelihood potentials, and how this can be enhanced for the many households that are not

able to do this, is an ongoing research challenge.


11

3.2 Asset accumulation and escapes from poverty

The process of accumulating key assets – or the fact of holding sufficient levels of assets to

begin with – are key factors enabling individuals and households to escape from poverty. But

it is equally clear that many are unable to participate in adequate levels of asset

accumulation, and this is an important part of understanding what underlies chronic poverty.

There is strong evidence that levels of asset holdings are strongly associated with level of

income. In a study of rural areas in four African countries, Ellis and Freeman (2004) find

asset accumulation as a gradual process of trading up activities, with acquisition of livestock

and an ability to access non-farm self-employment activities being particularly important. The

extent to which households are able to trade up assets is crucial to increasing welfare, and

this route was limited for many by various factors. In these environments, livestock was

critical, playing multiple roles in livelihood successes (e.g. savings, income smoothing); they

can form the basis of a virtuous circle for improving welfare, and their absence might

contribute to a poverty trap. In addition, Ellis and Freeman find that land productivity

increased sharply with income level (see discussion of asset-based poverty traps below).

Where households acquired access to even small plots of land in rural Mexico, this was often

found to raise household welfare substantially (Finan et al., 2005); but the value of the land

to the household depended also on their access to complementary assets, including having

primary education and access to a road. The importance of complementary factors also

comes out strongly in resettlement sites in rural Zimbabwe (Chimhowu, 2002). Acquiring

more land was important in enabling some households to escape from poverty, but for the

majority the land still gave inadequate income levels due to the absence of the other factors.

The key absent factors here appeared to be public policy interventions to provide social

services and infrastructure. In Southern Africa as a whole, Chimhowu (2006) argues that

land reform programmes have had limited poverty reduction impact (for various reasons);

and more generally Rigg (2006) also argues that land redistribution may not result in a

redistribution of wealth – partly because of the increasing importance of individuals engaging

in non-agricultural activities in order to escape poverty.

As this suggests, much of the literature on asset accumulation and escapes from poverty has

focused on agricultural environments; but some other studies have looked at the factors

influencing human capital, an important welfare outcome in its own right as well as a

determinant of other outcomes. In Latin America, Attanasio and Szekely (1999) highlight

many features of the close relationship between lack of education and poverty. In Costa Rica

and Peru poor people have lower levels of human capital, and earn lower returns on the

assets they do have in the labour market. In Brazil risk and uncertainly lead poor people to


12

invest in easily redeemed asset groups rather than human capital, and these assets are

much less productive in the longer term.

In environments where credit market imperfections limit the ability to invest in human capital

formation, migration as well as associated remittances can be important factors leading to

increased school attendance (Mansuri, 2006). Conditional cash transfers have played an

important role in leading to increased levels of secondary school enrolment and reduced

child labour in some Latin American countries ((MCDSS/GTZ, 2006; Haarman et al., 2008;

Samson et al., 2006; Rawlings and Rubio, 2005).

In many environments, access to wage employment is a key factor enabling escapes from

poverty (e.g. Whitehead, 2006; Krishna et al., 2006) human capital is obviously an important

enabler of this in many instances.

There is significant evidence therefore of the importance of asset accumulation in enabling

escapes from poverty, though at present the evidence base is strongest for rural areas (in

particular in relation to agriculture or movement out of agriculture) and in relation to human

capital. It is clear though that large numbers of individuals and households are unable to

accumulate key assets; and understanding why this is, and how policy might respond, is a

key research issue. The ways in which markets operate, as well as the nature of existing

public policy, are likely to be key dimensions of this.

3.3 Returns to assets and poverty transitions

As highlighted above, it is not just the level of assets to which individuals and households

have access that is important for their poverty status, but also the uses they can make of

these assets. Owning or having access to assets is of limited value if these assets cannot be

utilised effectively.

Rigg (2006) argues that land is no longer a prerequisite for enabling poor rural households to

escape poverty. He argues that rural livelihoods have become increasingly separated from

agricultural production; as is widely recognised, the importance of non-agricultural activities

has been increasing significantly over time and this may offer the key escape route from

poverty for many rural households. According to his analysis, the view that the redistribution

of resources will reinvigorate agricultural production is incorrect. This implies that the solution

to rural poverty lies within rural areas and within agriculture, a view that Rigg strongly

challenges (the ‘yeoman farmer fallacy’).

Traditionally, poverty has been associated with lack of integration into markets and lack of

state provision, especially in remote rural areas. Overcoming these potentially generates

increased opportunities to trade and so to raise incomes. But Rigg (2006) raises the issue of

whether this may be partly replaced by ‘new poverty’ associated with terms of integration into

markets. He questions the extent to which poor people in rural areas can benefit substantially


13

from opportunities offered by increased commodification and market integration. Whether

this is the case or not, it is clear that many of the best poverty reduction opportunities lie

outside of agriculture. As noted above, individuals frequently perceive the best option as

being a wage job outside the agricultural sector.

In rural Uganda important correlates of escape from poverty were increased productivity and

diversification into cash crops (Krishna et al., 2006). But again, only some households were

able to achieve this. Other asset-related factors associated with movements out of poverty

included successes in own business (a return to physical and/or human capital), or through

obtaining a private wage job – a factor which is repeatedly identified as an important factor

behind escapes from poverty in many countries.

In the case of North East Ghana, land was not identified as being a major constraint, but

rather labour and capital were the key factors (Whitehead, 2006). The social management of

household labour and household membership was critical for household security – in

particular ensuring that there was enough household labour supply available. There was

therefore a virtuous circle between household size (the amount of labour within a compound)

and wealth; and a poverty trap whereby households with too little labour had inadequate

wealth levels. Descents into poverty were often associated with temporary sale of labour

(with adverse consequences for farm production) or migration. These factors may also

translate into poverty traps. Again, access to secure employment was a key factor enabling

accumulation by households, enabling them to ‘straddle’ the urban and rural environment.

Lack of secure rural employment opportunities, however, was a major constraint for this

option.

The key question then is to understand the barriers to entry to the non-farm sector; is it just

level of assets (for example human capital) or do other factors matter (for example

discrimination or connections)? Ellis and Freeman (2004) similarly identify the importance of

reducing reliance on agriculture for escaping from poverty. Sen highlights for Bangladesh the

strong relationship between escapes from poverty and engaging in multiple livelihood

strategies. Frequently livelihoods may be unsustainable, as in the case of rickshaw pullers in

Dhaka, Bangladesh (Begum and Sen, 2004). For many this becomes a poverty trap, and is

also associated with intragenerational transmission of poverty in that the pullers, children, are

much less likely to attend school.

3.4 Other factors in relation to assets

One important issue of asset ownership to consider is who within the household controls the

assets to which it has access. Hoddinott (2006) finds important intra-household differences in

the extent of consumption smoothing in Zimbabwe. In Ghana Doss (2005) finds that

households where women have a higher share of asset ownership have better health and


14

nutritional outcomes, other things being equal; and Valdiva (2001) reports better welfare

outcomes in Indonesia when women household members control more assets.

In general it is clear that different individuals and households are able to attain different

returns on the assets to which they have access. Sometimes this reflects the level of the

assets – or the specific combinations of assets – to which they have access. It is equally

possible that two people or households with similar levels of asset holdings may earn

differential returns; this is frequently analysed in terms of discrimination. There have been

relatively few empirical studies (outside of the labour market) that have sought to analyse the

extent to which discrimination of this form may be an important factor underlying chronic

poverty.

Fundamentally what is still needed is a better understanding of the processes and factors

which underlie the differential returns individuals and households are able to earn on assets

they command.


15

4 Assets in relation to poverty dynamics

The conventional poverty dynamics approach distinguishes transient and chronic poverty,

which are typically identified using panel household survey data, and predominantly in the

monetary dimension (income/consumption; Hulme and McKay, 2007). Both spells and

components approaches have been used to measure chronic poverty (McKay and Lawson,

2003). In either of these approaches, however, the identification of people in chronic and

transient poverty may be significantly affected by measurement error; this is particularly so

for the spells approach, where the extent of transient poverty will tend to be overestimated,

due to measurement error. In addition, though, some transient poverty is also likely to be a

reflection of good or bad luck in one period. For these and other reasons, a promising

alternative approach may be to use data on household assets to distinguish between the

structurally poor and the stochastically poor (Carter and May, 1999, 2001). Structurally poor

people are those whose current asset levels are sufficiently low that they are unlikely to be

able to rise above the poverty line in future (without new asset accumulation in future).

Stochastically poor people are those who are poor in one or more periods, but whose asset

levels appear sufficient, suggesting that their poverty may reflect bad luck (adverse

stochastic factors) in one specific period, which may not have longer-term consequences.

Many households that are identified as chronically poor in income terms may be structurally

poor in asset terms; and likewise a persistently non-poor household might be expected to be

likely to be structurally non-poor. Transient poor households, though, may be stochastically

poor or non-poor: for example, the fact that a household was poor in the first period may just

have been a reflection of bad luck in that specific period. Or they may have made a structural

shift, as reflected in changing asset levels. The distinction between structurally and

stochastically poor households can be made based on a regression relationship between

household income and assets, which allows identification of an asset poverty line

corresponding to a given income poverty line. This approach assumes that assets can be

represented in a single dimensions, which is typically done by combining different

dimensions using factor analysis (Sahn and Stifel, 2000). There are, however, strong

assumptions in combining assets in this way.

For any one household it is possible to ask whether on average the level of assets it

accesses is sufficient to place it above the poverty line. If the level of assets in a period is

sufficient to place it above the poverty line, but the household is in fact income poor, then

that poverty can be considered as stochastic. If the level of assets is insufficient and the

household is poor, then that poverty can be represented as structural. A similar logic applies

to changes over time. If a household that escapes from poverty between two periods moves

from a situation where its first period assets were below the asset poverty line to having

asset levels above the asset poverty line in a second period, then this can be considered as

a structural escape from poverty. If another household that escaped from poverty had levels

of assets below the asset poverty line in both periods, then its escape from poverty can be


16

regarded as stochastic, reflecting a favourable but potentially one-off shock. In this way

poverty status at a point in time and poverty transitions (or non-transitions) between two

points in time can be classified as structural or stochastic.

To be able to make these judgements with confidence requires relatively comprehensive

information about assets, and trust in the factor analysis approach to combining different

assets. In practice, the asset index is usually constructed based on material assets only,

although sometimes human capital is also included. Various assumptions are made in this

process, including: (i) that similar assets are important for all households and using the same

weights; (ii) that all asset categories are adequately measured; and (iii) that the returns to

assets do not differ substantially across different households. All of these, and particularly

perhaps (i), are strong assumptions.

Figure 1: Assets and livelihood options

Source: Carter and Barrett (2006).

In the end, though, this approach is static. Carter and Barrett (2006) therefore develop a

dynamic asset threshold approach, focusing on asset accumulation over time. The key issue

is whether households face decreasing returns to scale in production throughout the entire

range, or whether there are areas of locally increasing returns to scale. Carter and Barrett

consider three possible causes of locally increasing returns:

(i) the underlying income-generating process has locally increasing returns;

(ii) a high return production process has a minimum effective size; or

(iii) marginal returns to wealth are lower for lower wealth households because of risk and

financial markets.


17

Taking the example of (ii) above, curve L2 in Figure 1 represents a high return activity which

however requires a minimum scale of operation. The diagram also shows the steady state

asset levels which would be adopted by two different households, one following strategy L1

(where the equilibrium asset level is A*1) and the other L2 (with asset level A*

2). The former

household, with few assets, cannot pursue the high return strategy and is consequently

caught in a poverty trap. The second household, with the higher level of assets, can pursue

the high return strategy and remains above the poverty line. Households with asset levels

below AS are required to pursue the low return strategy (and are condemned to a poverty

trap); those above AS will pursue the high return strategy. Households with assets between

A*1 and AS have low returns on their assets and have little incentive to save; they will

converge to the equilibrium at A*1.

It would of course be worth the households’ while to borrow to take their asset holdings up to

AS, which then would put them onto a path of accumulation (households with assets above

AS will converge to A*2); but the key point is that they are likely to be excluded from credit

markets and so many households below this level will be unable to borrow. Zimmerman and

Carter (2003) identify a Micawber threshold, as a level of assets above which it does make

sense to pursue an autarkic savings strategy (e.g. reducing consumption), which will

eventually enable the household to converge to the high outcome. If asset dynamics can be

represented as in the lower graph in Figure 2 below, then once households’ asset levels

reach the point A* (the dynamic asset or Micawber threshold), then they can accumulate

over time and this process of accumulation puts them on track to a point at which they

eventually switch to the high return activity L2. Below A*, though, they dissave and return to

the low equilibrium point. This constitutes a poverty trap from which households cannot

escape without external intervention. Thus there are both virtuous and vicious circles in asset

accumulation, depending on the level from which a household starts. In a situation with a

dynamic asset threshold there is a natural tendency to a divergence of living standards

between households over time, and thus a substantial risk that (in a growth environment at

least) inequality will tend to rise in the absence of external intervention.

This diagram was drawn for the case A*1< A* < A . A similar analysis applies if A*

1< A < A*

except that in the latter case not only the currently structurally poor households but some not

currently structurally poor households will tend to the low equilibrium.

Figure 2: Asset dynamics and the dynamic asset poverty line


18

An extension of the poverty trap concept by Barrett and Swallow (2006) introduces the

concept of fractal poverty traps. Analogous concepts of poverty traps can also apply at

community, regional and national levels; they refer to the situation where poverty traps exist

simultaneously at different levels (micro, meso and/or macro) and reinforce each other as a

fractal poverty trap. For example, governments, communities, markets and households may

all simultaneously find themselves in low level equilibria, making it very difficult to escape

from persistent poverty. In such an environment, it is very difficult to escape from past

disadvantage or adverse shocks.

If there is evidence for the existence of poverty traps, then this has important policy

implications for dealing with persistent poverty. In particular, while safety nets to prevent the

use of coping strategies which destroy assets are needed, these need to be complemented


19

by ‘cargo nets’ to enable households to build up a sufficient asset base to prevent poverty-

perpetuating practices. In different contexts this may imply some or all or measures which

help raise productivity of assets (e.g. improved market access; improved access to finance;

investment in education; and land reform). Carter and Barrett (2006) argue that the severity

of shocks households face may be less important for persistent poverty than the household’s

ability to manage these shocks through having a sufficient level of assets (relating to the

discussion of vulnerability in Section 2).

But before jumping to policy conclusions, it is necessary to understand when and if such

poverty traps and divergent processes may come about (and where they do, what the

fundamental cause is). The key question is the dynamics of asset accumulation and the

existence, at least over a range, of increasing returns to scale. If this range of increasing

returns to scale does not exist (if the asset accumulation curve does not have the S-shape

shown in the lower chart in Figure 2), then there is no rationale to support a poverty trap. If

this is the appropriate characterisation, then the poor are not prevented from escaping from

poverty by being trapped if they are in an environment of sustained growth. That said though,

it is likely to take them a very long time to escape, and they are still likely to need significant

policy intervention.

How strong is the empirical evidence in support of poverty traps? The key empirical question

is to find evidence of non-linearity in the accumulation relationship (i.e. the analogue of

Figure 2 above). Lybbert et al. (2004) estimate livestock dynamics in a pastoral population –

where there is only one core asset, livestock – and do report evidence of this non-linear

pattern. Barrett et al. (2006) find similar evidence in rural Kenya and Madagascar, while

Adato et al. (2006) find evidence for dynamic asset thresholds in environments where

multiple assets are relevant. The econometric approaches adopted in these studies, though,

do depend on strong econometric assumptions – in particular, the assumption that the

process of accumulation for all households can be represented on the same curve, which is

hard to justify. In addition, the dynamic poverty trap idea concerns the likely trajectory of a

household over time, but is estimated based on a cross section of households at a specific

point in time.

Quite a few recent studies do not find evidence in support of poverty traps. In multi-asset

settings Naschold (2008) finds evidence from Ethiopia, India and Pakistan for a weakly

concave dynamic asset threshold, in other words evidence in support of (very slow)

convergence. Baulch and Quisumbing (2009) also do not find evidence in support of poverty

traps in Bangladesh, despite having tried many specifications, and nor does Perge (2010) in

a forest community in Bolivia. In earlier studies, Jalan and Ravallion (2004) and Lokshin and

Ravallion (2004) also do not find evidence for multiple dynamic equilibria. Thus the case for

poverty traps has mostly been established in cases where households primarily rely on one

asset; in a broader context, where more assets are relevant, the evidence in favour of


20

poverty traps is much less convincing. Of course this is completely consistent with the view

that it may take a very long time for the poor to escape from poverty.

The ideas about poverty traps developed by Carter, Barrett and others capture a strongly

intuitive notion – that many individuals or households may be trapped in persistent poverty

reflecting a lack of assets combined with disadvantaged access to credit and financial

markets, or to insurance. But in the end the empirical evidence in support of this notion is not

that strong.


21

5 Discrimination as a factor in access to and use of assets

The logic of the poverty traps approach is that many households are caught in persistent

poverty because they do not have a sufficient level of assets. There is little question that lack

of assets, if sustained over time, is likely to be a major factor underlying long-term poverty.

Further, there is very strong evidence for those lacking sufficient assets being excluded from

credit markets (Payne et al., 2007); and financial markets in general, and it is widely

reported that even microcredit programmes which in principle target the poor frequently fail to

reach the poorest (Hulme, 2000; Kabeer, 2008; Mosley and Hulme, 2009)

But the poverty trap framework (at least as developed above) does not entertain the

possibility that one household may get a higher return on the same assets than another. This

of course can reflect good or bad luck, which may average out over a sufficiently long period

of time. But, equally, it may reflect discrimination in the way in which households are able to

make use of the assets they have. Some may systematically receive lower returns on assets

than others. Furthermore, there may also be discrimination in asset accumulation

opportunities in the first place.

Discrimination may take different forms – common examples being discrimination by gender,

ethnicity, caste, religion, geographic location or income level. Some of this may be about

‘natural’ disadvantage of particular locations, e.g. remote areas will often be far from many

important markets, though of course there may be endogenous factors underlying

remoteness. It may also be about policy choice, e.g. to construct a new road to region X

rather than region Y. But in many instances discrimination is much more overt.

The classic analysis of discrimination in relation to the labour market obviously concerns the

rate of return to human capital. Gender and racial pay gaps have been extensively studied in

many countries, both industrialised and developing, with the classic approach being to use

wage regressions to separate out wage differentials between groups into elements that can

be explained in terms of the characteristics of the groups, e.g. their respective education

levels, and that which cannot be explained by difference in characteristics. The latter is

interpreted as discrimination.

Besides discrimination in wage rates paid, there is also potentially discrimination in

accessing jobs in the first place, or discrimination in sectoral allocation of jobs (presumably

limiting access to better paid jobs). In practice, these factors might be more important than

wage differentials; in many countries getting a waged job (outside of agriculture) is a key

factor enabling an escape from poverty. This issue has been the subject of many studies,

which typically involve modelling the likelihood of getting a job separately for the advantaged

and disadvantaged group(s) as a function of their characteristics; or modelling sectoral

allocation separately for the advantaged and disadvantaged groups. Again, from these


22

studies there is extensive evidence of discriminated groups suffering from disadvantage. In

the Indian context, Thorat and Mahamallik (2007) provides a summary of evidence on

discrimination in the labour market against scheduled castes, and finds evidence that it is

substantial in hiring for jobs as well as wage payments and working conditions (despite

affirmative action measures at least in the public sector).

But the differentials in characteristics – in particular education – which play a central role in

these models in turn may themselves be the outcome of a discriminatory process:

discriminated groups may have greater difficulty in accessing education in the first place.

This concerns inequality, and possible discrimination, in accumulating assets in the first

place. This is a key starting point for discrimination. Again there is a significant econometric

literature on differentials in access to education (enrolment, attainment) by different groups,

in particular in relation to gender. In this context there is a specific interest in disadvantage

faced specifically by poor people in discriminated groups; reliance on results derived from

quantile regression analysis (where the line is calculated not relative to the mean value of the

variables, but about specific quantiles, including lower ones, of the distribution) may be

particularly informative. One factor which is generally relevant in models of child educational

outcomes is the education level of the children’s parents, which itself is likely to have been

an outcome of a discriminatory process in the past.

But the issue of discrimination is also important in other contexts besides the labour market.

Discrimination may be active in terms of access to land, access to markets, or access to

other public facilities, such as healthcare. Thorat and Mahamallik (2007) in India reviews

discrimination faced by scheduled castes and policy to remedy this in relation to the land

market. Scheduled castes (SCs) and tribes also suffer from lack of access to markets for

consumer goods and agricultural products, among other things; and traditionally were

debarred from undertaking any business. SCs face disadvantage in access to credit for

reasons that do not just reflect lack of assets. Many of these forms of discrimination have

become less important over time, though they still seem to exist on a substantial scale.

Discrimination in access to inputs, markets or public services also has important implications

for accumulation of assets and returns to assets, just as was the case for the labour market,

and in many countries these forms of discrimination have more severe consequences for the

chronically poor and the poorest.

Discrimination is likely to have serious adverse consequences for significant numbers of the

chronically poor, who may suffer from it and may be poor as a direct result. The quantitative

contribution of discrimination to chronic poverty has never been satisfactorily established. In

part, this is because the above techniques cannot in the end distinguish discrimination from

other observed differences between advantaged and disadvantaged groups. It is also partly

because discrimination may arise in many different spheres. In addition, analysis has

focused predominantly on the labour market, which is only one aspect. But in some low

income countries at least, the impact of discrimination on chronic poverty is likely to be high.


23

But in addition, there is increasing evidence that discrimination may also have adverse

impacts on overall economic performance. This evidence is probably strongest in relation to

gender inequality, which appears to have adverse implications for economic growth. Thus,

measures to tackle discrimination are likely to be good for chronically poor people, both

directly and indirectly, through enabling an improved overall economic performance.

In summary, discrimination in access to assets and in the ability to use them is potentially an

important factor underlying chronic poverty. There is scope to analyse this much more fully in

relation to markets, in addition to that for labour.


24

6 Conclusions

Assets play a key role in the understanding of poverty, and in particular in relation to chronic

poverty. First, it is important to acknowledge that poor people, including the chronically poor,

do have access to a number of assets, and this needs to be recognised in strategies to seek

to escape from poverty. But, that said, poor people, almost by definition, have relatively low

levels of assets; or do not have a broad enough range of assets. In addition, they are likely to

be able to make less effective use of the assets they do have, for various reasons, including

the very fact that their ownership is low or does not cover a wide enough range of assets;

that they may face other disadvantages (such as being in a remote location); or that they

suffer from discrimination or political exclusion.

Low levels of assets equate to higher levels of vulnerability. As households will seek to avoid

having to sell the few assets they may have, this will typically lead to more risk-averse

behaviour, in terms of choice of economic activities, which in itself may keep households

poor. One particular form of this is models of poverty traps, where lack of access to credit

accentuates the disadvantages poor people face due to low levels of assets. Empirical

evidence for poverty traps may not be very strong, but there are likely to be clear forces at

work making it very difficult for the chronically asset poor to escape poverty. This may be

further exacerbated in some cases by other disadvantage, discrimination or exclusion. These

latter factors, combined with low asset levels and limited opportunities to use these assets,

are likely to be very important factors in keeping people in chronic poverty. On the other

hand, having a sufficient level of assets in the first place is an important factor enabling

households to participate in processes of economic growth, as well as to protect themselves

against downturns.

There is extensive evidence that loss of assets, often through health shocks, is an important

factor underlying descents into poverty, and of course this is more likely for those with fewer

assets to begin with. There is some weaker evidence that accumulation of assets can enable

escapes from poverty, but this is often a long-term and gradual process.

In terms of a research agenda, the relationship between assets and vulnerability, and the

role of asset loss in descents into poverty are both very well established. There is, however,

scope for further research to investigate the possibility of poverty traps, and for a much more

thorough analysis of discrimination, in particular outside labour markets, where the

phenomenon is relatively well studied. Discrimination is one factor potentially underling

persistence of asset inequality over time, a factor likely to be strongly associated with chronic

poverty. It is also relatively well known that the extent of asset ownership is associated with

power, which becomes an important factor behind the persistence of a status quo in which

many remain poor. There is, however, scope for further analysis of this issue, in particular to

see to what extent initiatives to build up the social capital of poor people might help to

challenge this. In addition, it is important to know more about the processes of asset


25

accumulation and asset transactions, the latter in particular through markets and inheritance.

These latter two factors may be important factors behind the persistence of asset inequality

(and so of poverty), including its transmission from one generation to another.


26

References

Adato, M.; Carter, M. and May, J. (2006). ‘Exploring poverty traps and social exclusion in South Africa using qualitative and quantitative data’. Journal of Development Studies, 42(2), 226-247.

Ahmed, S. M. (2005). Healthcare for the Ultra-poor: The Case of BRAC’s CFPR/TUP Programme. London: Department for International Development Health Systems Resource Centre (DFID HSRC).

Arun, S. (2008). ‘Managing assets and vulnerability contexts: vistas of gendered

livelihoods of adivasi women in South India’. Brooks World Poverty Institute (BWPI)

Working Paper No. 32. Manchester, UK: The University of Manchester.

Attanasio, O. and Szekely, M. (1999). ‘An Asset-Based Approach to the Analysis of Poverty in Latin America’, Working Paper R-376. Washington, DC: Inter-American Development Bank.

Barrett, C. (1999). ‘Food Security and Food Assistance Programs’. In Bruce L. Gardner

and Gordon C. Rausser (eds), Handbook of Agricultural Economics. Amsterdam: Elsevier Science.

Barrett, C., Carter, M. and Little, P. (2006a). ‘Understanding and reducing persistent poverty in Africa: introduction to a special issue’. Journal of Development Studies, 42(2), 167–77.

Barrett, C., Marenya, P., McPeak, J., Minten, B., Murithi, F., Oluoch-Kosura, W., Place, F., Randrianarisoa, J., Rasambainarivo, J., and Wangila, J. (2006b). ‘Welfare dynamics in rural Kenya and Madagascar’. Journal of Development Studies, 42(2), 248–77.

Barrett, C.B. and Swallow, B.M. (2006).’Fractal Poverty Traps. World Development, 34(1), 1-15.

Baulch, B. And Quisumbing, A. (2009). ‘Assets and Poverty Traps in Rural Bangladesh’. CPRC Working Paper No. 143. Manchester, UK: Chronic Poverty Research Centre.

Begum, S. and Sen, B. (2004). ‘Unsustainable Livelihoods, Health Shocks and Urban Chronic Poverty: Rickshaw Pullers as a Case Study’. Chronic Poverty Research Centre Working Paper No. 46. Manchester, UK: Chronic Poverty Research Centre (CPRC).

Carter, M. and Barrett, C. (2006). ‘The economics of poverty traps and persistent poverty: an asset-based approach’. Journal of Development Studies, 42(2), 178–99.

Carter, M.R. and May, J. (1999). ‘Poverty, Livelihood and Class in Rural South Africa. World Development 27(1), 1-20.

Carter, M. and May, J. (2001). ‘One kind of freedom: the dynamics of poverty in post-apartheid South Africa’. World Development, 29, 1987–2006.

Chambers, R. (2003). ‘Vulnerability, coping and policy’, IDS Bulletin 37(4): 33-40.

Chimhowu, A. (2006). ‘Tinkering on the Fringes? Redistributive Land Reforms and Chronic Poverty in Southern Africa’. Chronic Poverty Research Centre Working Paper No. 58. Manchester, UK: Chronic Poverty Research Centre (CPRC).


27

Chimhowu, A. (2002). ‘Extending the grain basket to the margins: spontaneous land resettlement and changing livelihoods in the Hurungwe District, Zimbabwe’.Journal of Southern African Studies, 28(3).

Corbett, J. (1988). ‘Famine and Coping Strategies’. World Development, 16(9), 1099-1112.

Dercon, S. (2000). ‘Income risk, coping strategies and safety nets’. CSAE Working Paper 136. Oxford: The Centre for the Study of African Economies.

Devereux (1993). ‘Goats before ploughs: Dilemmas of household response sequencing during food shortage’. Institute of Development Studies Bulletin, 24 (4), 52-59.

Devereux, S. (2002). ‘Livelihood insecurity and social protection: a re-emerging issue in rural development’. Development Policy Review 19(4): 507-519.

Doss, C. (2005). ‘The effects of intrahousehold property ownership on expenditure patterns in Ghana’. Journal of African Economies, 15(1), 149–80.

Ellis, F. and Freeman, A. (2004). ‘Rural livelihoods and poverty reduction strategies in four African countries’. Journal of Development Studies, 40(4), 1–30.

Escobal, J., Saavedra, J. and Suarez, P. (2005). ‘Economic Shocks and Changes in School Attendance and Levels and Education Expenditure in Peru’. Young Lives Working Paper No. 13. Oxford: Young Lives, An International Study of Childhood Poverty.

Fafchamps, M.; Udry, C. and Czukas, K. (1998). ‘Drought and Saving in West Africa: are Livestock a Buffer Stock?’. Journal of Development Economics, 55(2), 273-305.

Finan, F., Sadoulet, E. and de Janvry, A. (2005). ‘Measuring the Poverty Reduction Potential of Land in Rural Mexico.’ Journal of Development Economics, 77(1), 27-51.

González de la Rocha, M. (2006). ‘Vanishing assets: cumulative disadvantage among the urban poor’. The Annals of the American Academy of Political and Social Science, 606(1): 68-94.

González de la Rocha, M. (2007). ‘The construction of the myth of survival’, Development and Change, 38(1): 45-66.

Haarman, C., Haarman, D., Jauch, H., Mote, S., Nattrass, Samson, M., and Standing, G. (2008). ‘Towards a basic income grant for all: basic income grant pilot project assessment report’. Namibia: Namibia Non-Governmental Forum (NANGOF).

Hamoudi, A. and Birdsall, N. (2001). ‘AIDS and the Accumulation and Utilization of Human Capital in Africa’. Center for Global Development, Working Paper No. 2. Washington, DC: Center for Global Development.

Harpham, T. and Grant, E. (2002). ‘Health, health services and environmental health’. In Rakodi, C. and Lloyd-Jones, T. (eds.) Urban livelihoods: a people-centred approach to reducing poverty. London: Earthscan.

Hoddinott, J. (2006). ‘Shocks and their consequences across and within households in rural Zimbabwe’. Journal of Development Studies, 42(2), 301–21.

Howe, G. and A. McKay. (2007). ‘Combining Quantitative and Qualitative Methods in Assessing Chronic Poverty: The Case of Rwanda’. World Development, 35(2), 197-211.


28

Hulme, D. and A. McKay (2007). ‘Identifying and Measuring Chronic Poverty: Beyond Monetary Measures?’. In N. Kakwani and J. Silber (eds.), The Many Dimensions of Poverty. Hampshire: Palgrave-Macmillan.

Hulme, D. (2000) ‘Is microdebt good for poor people? A note on the dark side of microfinance’. Small Enterprise Development, 11(1): 26-28.

Jalan, J. and Ravallion, M. (2004). ‘Household income dynamics in rural China’. In Dercon, S. (Ed.), Insurance Against Poverty. Oxford: Oxford University Press.

Kabeer, N. (2008) ‘Mainstreaming gender in social protection for the informal economy’. London: Commonwealth Secretariat.

Krishna, A. (2004). ‘Escaping poverty and becoming poor: who Gains, who loses, and why?’. World Development, 32(1), 121–36.

Krishna, A. et al. (2006). ‘Escaping poverty and becoming poor in 36 villages of Central and Western Uganda’. Journal of Development Studies, 42(2), 346–70.

Lokshim, M. and Ravallion, M. (2004). ‘Household Income Dynamics in Two

Transition Economies’. Studies in Nonlinear Dynamics and Econometrics, 8(3), Article 4.

Lourenço-Lindell, I. (2002). ‘Walking the Tight Rope: Informal Livelihoods and Social Networks in a West African City’. Stockholm: Stockholm Studies in Human Geography 9, Almqvist and Wiksell International.

Lybbert, T.J., Barrett, C.B., Desta, S. and Coppock, D.L. (2004). ‘Stochastic Wealth Dynamics and Risk Management Among a Poor Population’. Economic Journal, 114(498), 750-777.

Mansuri, G. (2006). ‘Migration, School Attainment and Child Labor: Evidence from Rural Pakistan’.World Bank Policy Research Working Paper 3945. Washington, DC: Development Research Group, The World Bank.

Maxwell, S. and Smith, M. (1992). ‘Household Food Security: A Conceptual Review’. In S. Maxwell and T. Frankenberger (eds), Household food Security: Concepts, Indicators, and Measurements: A Technical Review. New York and Rome: UNICEF and IFAD.

McKay, A. and D. Lawson (2003). ‘Assessing the Extent and Nature of Chronic Poverty in Low Income Countries: Issues and Evidence.’ World Development 31(3): 425-439.

Ministry of Community Development and Social Services (MCDSS) and German Technical Cooperation (GTZ) (2006). Evaluation Report –Kalomo Social Cash Transfers Scheme. Lusaka:MCDSS/GTZ.

Mitlin, D. (2003). ‘Addressing urban poverty through strengthening assets’, Habitat International, 27: 393-406.

Moser, C. (1998). ‘The asset vulnerability framework: reassessing urban poverty reduction strategies’. World Development, 26(1): 1-19.

Mosley, P. and Hulme, D. (2009) ‘Microenterprise finance: is there a conflict between growth and poverty alleviation?’. In Hulme, D. and Arun, T. (eds.) Microfinance: a reader. Abingdon: Routledge.


29

Naschold, F. (2008). ‘Four Papers on Structural Household Welfare Dynamics’. PhD thesis, Cornell University.

Nkurunziza, E. and Rakodi, C. (2005). ‘Urban families under pressure: conceptual and methodological issues in the study of poverty, HIV/AIDS and livelihood strategies’. Working Paper 1. Birmingham: University of Birmingham.

Payne, G., Durand-Lasserve, A. and Rakodi, C. (2007) ‘Social and economic impacts of land titling programmes in urban and peri-urban areas: a review of the literature’. Presented at the World Bank Urban Research Symposium, Washington, D.C., 14-16 May 2007.

Perge, E. (2010) ‘Testing a Poverty Trap Mechanism with Tsimane’ Panel Data’. CPRC working paper No. 158. Manchester, UK: Chronic Poverty Research Centre.

Rawlings, L. and Rubio, G. (2005). ‘Evaluating the Impact of Cash Transfer Programs’. World Bank Policy Research Working Paper 3119. Washington, DC: Human Development Sector Unit, Latin America and the Caribbean Region, The World Bank.

Rigg, J. (2006). ‘Land, farming, livelihoods, and poverty: rethinking the links in the rural South’. World Development, 34(1), 180–202.

Sahn, D.E. and Stifel, D.C. (2000). ‘Poverty Comparisons Over Time and Across Countries in Africa’. World Development, 28(12), 2123-2155.

Samson, M., van Niekerk, I. and MacQuene, K. (2006). ‘Designing and Implementing Social Transfer Programmes’. Cape Town: Economic and Policy Research Institute Press (EPRI).

Sen, A. (1991). ‘Capability and Well-Being’. In M. Nussbaum and A. Sen (eds), The Quality of Life. Oxford: Clarendon Press.

Sen, B. (2004). ‘Drivers of Escape and Descent: Changing Household Fortunes in Rural Bangladesh’, Bangladesh Institute of Development Studies. Available at: http://www.prcpb-bids.org/documents/Drivers%20of%20Escape_bSen.PDF

Swift (1989). Why are Rural People Vulnerable to Famine?’. IDS Bulletin, 20(2): 8–15.

Thorat, S. and Mahamallik, M. (2007) ‘Chronic poverty and socially disadvantaged groups: analysis of causes and remedies’. CPRC-IIPA Working Paper No. 33. Chronic Poverty Research Centre/Indian Institute of Public Administration.

Valdiva, C. (2001). ‘Gender, livestock assets, resource management, and food security: lessons from the SR-CRSP’. Agriculture and Human Values, 18, 27–39.

Whitehead, A. (2006).‘Persistent poverty in North East Ghana’. Journal of Development Studies, 42(2), 278–300.

Woolcock, M. (2005). ‘Calling on friends and relatives: social capital’. In Fay, M. (ed.) The urban poor in Latin America. Washington, D.C.: The World Bank.

Zimmerman, F. J. and Carter, M. R. (2003). ‘Asset smoothing, consumption smoothing and the reproduction of inequality under risk and subsistence constraints’. Journal of Development Economics, 71, 233–60.

http://www.prcpb-bids.org/documents/Drivers%20of%20Escape_bSen.PDF

http://www.prcpb-bids.org/documents/Drivers%20of%20Escape_bSen.PDF

The Chronic Poverty Research Centre (CPRC) is an international partnership of universities, research institutes and NGOs, with the central aim of creating knowledge that contributes to both the speed and quality of poverty reduction, and a focus on assisting those who are trapped in poverty, particularly in sub-Saharan Africa and South Asia.

Partners: Bangladesh Institute of Development Studies (BIDS), Bangladesh Development Initiatives, UK Development Research and Training, Uganda Economic Policy Research Center, Uganda FIDESPRA, Benin HelpAge International, UK Indian Institute of Public Administration, India IED Afrique, Senegal Institute of Development Studies, UK Overseas Development Institute, UK Programme for Land and Agrarian Studies, South Africa University of Legon, Ghana University of Manchester, UK University of Sussex, UK

Contact: [email protected] © Chronic Poverty Research Centre 2009

www.chronicpoverty.org

The Intergenerational Transmission of Poverty during the AIDS Epidemic … · 2016-08-02 · He has...

Documents

Transcript of The Intergenerational Transmission of Poverty during the AIDS Epidemic … · 2016-08-02 · He has...